Document management that doesn’t suck


Props to @MylesCarrick for the title. He sparked a conversation on twitter this week with that sentiment. So, I’m wondering what you look for in a document management system.

Here’s a few things I can think of:

1. Some sort of built-in, configurable governance for file naming convention—A lot of the problem I have with document management systems is that people still name things randomly and folders are often filled with unrelated, random contents. Some guidance for naming convention that didn’t rely on verbal reinforcement would be ideal.

2. Files and folders default to public. Explicit exclusion for folders/files that must be private—One of the challenges of knowledge management is that, as an organisation, we don’t know what we know. Transparent file storage allows for discoverability when we’re searching for something. The current sharing models require us to explicitly ALLOW access, rather than explicitly DENY.

3. Semantic clustering and recommendation engines—Related to my last post about miscellany, I’d like to see a DMS that can offer us suggestions of similar documents based on keywords and phrases. Of course, this would be dependent on point 2, above.

Are you happy with your document management solution? What are the things you look for?

  • JGander

    A good simple way of searching for knowledge articles that sorts the documents in usefulness order. Nothing worse than having to search through search results to find what you need.

    Simple tagging approach that doesn’t require additional time to add / create a new tag, and advises you if there is a similar tag in use already.

  • NateBeran

    We’ve been going through an enterprise document management implementation project for the last 3 months. This isn’t the first time I’ve gone through one of these but its the first time the vendor really sat down and worked on document classification with my team. Since they’re focused on digitizing docs (old school!) they have a distinct vision for how documents get classified and indexed so that you can retrieve them in a meaningful manner later. Custom searches and search forms become really important as your system grows. Nothing goes into the system without index data and a document type assignment, so the administrator team has some powerful data to build queries out of.

    That’s all great for “documents” but I still feel that the traditional “document” is the worst possible way to store data. Especially when you want to facilitate knowledge transfer. I’m still searching for the best ways to do that. Blogs, wikis, and many of the other ajax-y and web 2.0 tools seem pretty good but I’m still searching.